powerpc/booke: Only check for hugetlb in flush if vma != NULL
authorScott Wood <scottwood@freescale.com>
Fri, 22 Nov 2013 00:26:42 +0000 (18:26 -0600)
committerScott Wood <scottwood@freescale.com>
Fri, 22 Nov 2013 22:57:29 +0000 (16:57 -0600)
And in flush_hugetlb_page(), don't check whether vma is NULL after
we've already dereferenced it.

This was found by Dan using static analysis as described here:
https://lists.ozlabs.org/pipermail/linuxppc-dev/2013-November/113161.html

We currently get away with this because the callers that currently pass
NULL for vma seem to be 32-bit-only (e.g. highmem, and
CONFIG_DEBUG_PGALLOC in pgtable_32.c)  Hugetlb is currently 64-bit only,
so we never saw a NULL vma here.

Signed-off-by: Scott Wood <scottwood@freescale.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
arch/powerpc/mm/hugetlbpage-book3e.c
arch/powerpc/mm/tlb_nohash.c

index 3bc7006..74551b5 100644 (file)
@@ -117,6 +117,5 @@ void flush_hugetlb_page(struct vm_area_struct *vma, unsigned long vmaddr)
        struct hstate *hstate = hstate_file(vma->vm_file);
        unsigned long tsize = huge_page_shift(hstate) - 10;
 
-       __flush_tlb_page(vma ? vma->vm_mm : NULL, vmaddr, tsize, 0);
-
+       __flush_tlb_page(vma->vm_mm, vmaddr, tsize, 0);
 }
index 41cd68d..358d743 100644 (file)
@@ -305,7 +305,7 @@ void __flush_tlb_page(struct mm_struct *mm, unsigned long vmaddr,
 void flush_tlb_page(struct vm_area_struct *vma, unsigned long vmaddr)
 {
 #ifdef CONFIG_HUGETLB_PAGE
-       if (is_vm_hugetlb_page(vma))
+       if (vma && is_vm_hugetlb_page(vma))
                flush_hugetlb_page(vma, vmaddr);
 #endif